Recently, a UNIX (registered trademark) server and an IA server have been introduced to a core system, and the high availability of the UNIX (registered trademark) server and the IA server is emphasized. Generally, when there occurs a fatal error in a system, the system is stopped in an emergency, and the memory dump is stored on the disc for search of the cause of the error.
While stopping the system, the system is not used. Therefore, it is very important to quickly restart the system.
However, in recent years, a server implemented with memory of a capacity in the terabyte (TB) order has been introduced. In such a system, it takes a long time to extract a memory dump, and a system is not quickly restarted.
There are also a method of storing memory contents at an emergency stop on another memory unit without storing a memory dump on a disc, and a method of storing apart of memory and converting the memory contents not stored after a restart into a dump file.
However, in the conventional methods, a memory dump at the occurrence of an abnormal condition is stored on another memory unit or disc. Therefore, if a large size of memory dump is to be stored, it takes a long time to copy the memory, thereby disabling the system to be quickly restarted.
When an operating system detects a fatal error and the system is stopped in an emergency, the operating system which has detected an abnormal condition extracts a dump. Therefore, the abnormal condition is detected again while dumping memory, thereby causing a secondary damage such as an occurrence of a hang-up etc.